Artificial intelligence (AI) and machine-learning (ML) have quickly grown beyond a few major tech companies and hardcore academic researchers. Every marketing organization can tap into the power of AI to streamline operations and grow the business.
The new book The AI Marketing Canvas: A Five-Stage Road Map to Implementing Artificial Intelligence in Marketing provides a growth framework for business and marketing leaders to implement AI using a five-stage model called the “AI Marketing Canvas.”
On this episode of Marketing Smarts, I speak with co-author Rajkumar Venkatesan about how he and his co-writer developed those stages by studying leading global brands. We cover examples of brands―including Google, Lyft and Coca-Cola―that have successfully woven AI into their marketing strategies.
This is not a conversation about coding for AI models. Raj and I talk about how marketing leaders can go from “zero to hero” with AI in marketing, and what that means for your team and your company culture.
Listen to the entire show now from the link above, or download the mp3 and listen at your convenience. Of course, you can also subscribe to the Marketing Smarts podcast in iTunes or via RSS and never miss an episode.
"Marketing Smarts" theme music composed by Juanito Pascual of Signature Tones.
Kerry O'Shea Gorgone: Welcome to the Marketing Smarts Podcast. I'm here with Raj Venkatesan, co-author of the book The AI Marketing Canvas: A Five Stage Roadmap for Implementing Artificial Intelligence in Marketing. He's the Ronald Trzcinski professor of business administration in the Darden Graduate School of Business Administration in Virginia. His writing has appeared in The Journal of Marketing and The Harvard Business Review. Let's just say he's very smart and knows a lot of things about a lot of things.
How did you realize, Raj, that AI was going to be powerful for marketing specifically? When you learned it, did you learn it in the context of marketing, or did you learn it in some other context first?
Rajkumar Venkatesan: Thank you, Kerry, for having me. It's a pleasure to be here. Thank you for your kind introduction. The story about how I started working on AI in marketing, it's a really interesting question you ask about how I learned about AI first, it goes back to the mid-90s when I was doing my undergrad in computer engineering. In my final year, I learned about neural networks, and my final year project was on genetic algorithms. Those are some of the things that are in the AI toolkit now. I started there.
When I did my PhD, my advisor recognized that computers and technology are going to shape marketing in the future and asked if I wanted to work on that. I didn't know much about how things were going to turn out, and I said I'd do that. That's how I started. So, I came into this as a technology computer engineer, but as I started taking classes in economics, marketing, and consumer behavior, I started realizing it was very fascinating to hear about all of these new subjects and also it gave me a sense of where these tools that I was learning could be applied and how they can actually be valuable.
I pursued and I started teaching, I came to UVA in 2006 after teaching in University of Connecticut. I started teaching marketing analytics. Darden is a special case where we value practice and connection with a practicing manager and we do the case method, which really teaches me about marketing and also understanding how managers are using data and what challenges they're solving every day. As I was teaching my course, and writing my case studies, and doing my research, I started seeing that more and more you now have data and technology influencing marketing.
That's how the worlds kind of came together for me and I started working on this book.
Kerry: I think a lot of us as marketers probably see the real world results of AI kind of at work. Like Starbucks knows I always get this certain thing and it will show me the thing every time I open the app, and it will suggest other things based on my history and that kind of thing. I can see that in the B-to-C context really clearly. What does AI look like for B-to-B marketers?
Raj: Great question. In fact, when I started working on data in marketing and I did my work on customer lifetime value, it was in the context of B-to-B. I think one of the things with B-to-B which is really an advantage is that they have a direct relationship with the customers most of the time. Of course, they can sell through resellers, but to a large extent B-to-B firms know their customers.
With Salesforce and other software today, there is a lot of information about how salespeople are connecting with their customers and what the customers are talking about. Of course, there are challenges, like salespeople actually putting the data in and how much is the coverage and all of that.
I think where it can be really useful is the fact that you know about who your customers are and what they are buying, and you know a lot more detail about your customers in terms of how they're using your product because of your client relationship specialists and managers. You're really embedded with your customers, which really gives you information about how they are using your products. With Cloud and IoT, there's a lot more information coming these days about usage.
Kerry: And lack of usage.
Raj: Or lack of usage, yes. I think the potential is there. A lot of potential is there for B-to-B in actually doing sales call planning, whys of customer, coming up with new product updates, enabling a robust customer community through educational videos, FAQs, or even understanding if a customer is coming to a webinar, or if they ask some questions in the webinar, or have been to your website, browsed on the website. Information about that is really useful for salespeople when they go and talk to the customer after that in a follow up call, they can start becoming relevant and personal.
I think B-to-B is definitely transforming. There's a lot more interaction, like web 2.0 and whitepapers and content marketing, and all of that is rich with data and are ripe, fertile areas for B-to-B marketers to use data to better improve the effectiveness of their marketing activities.
Kerry: So, you could create a B-to-B content hub and then analyze the heck out of people's interaction with it?
Raj: Yes, absolutely. Think about if you have a product, let's say you're Siemens and you're making parts, maybe you're making engines for trains or aircraft, as an example. There are many different parts that go into it and you're also supplying the parts.
Knowing what kind of whitepapers your customer is downloading can tell you whether they're interested in maintenance, or whether they're interested in customizing your product, or understanding how your part fits within their project. It really helps your salespeople to then really hone in on that when they go on that follow up call. Also, on the emails you send them you can really start personalizing how you present your company to the customers.
Kerry: Can you talk to me about some of the research that went into the book?
Raj: With my co-author Jim Lecinski, I think we all came in with our own perspectives. He worked in Google for a while before he joined Northwestern. I was teaching marketing analytics and I started teaching digital marketing in a course called Marketing Technology Products where I would spend a week in San Francisco with my students talking, listening, and visiting companies. It gave me a look into how the world is evolving, what new technologies are coming.
When we put together all of this, we started out with here's a dump of what we think we know, and then we started talking to brands who were really in this journey. All of the brands are featured in our book from Washington Post, Coca-Cola, Unilever, Ancestry.com, Carmax and several others. It started giving us a picture about in-depth interviews with managers who are really trying to bring data into their marketing. It gives us a consistent pattern on how they go about building this AI capability in their functions.
That's kind of what led to us coming up with this canvas on where to begin and what the steps are to gain this transformative capability for their organizations.
Kerry: Not to give away the store, because we want everybody to buy the book, obviously, but can you talk about the five-stage roadmap at a high level?
Raj: Absolutely. I'm happy to. The five steps are foundation, experimentation, expansion, transformation, and monetization. What we mean by that is for AI to work, the raw material is data. But it's not any data, it is data that is focused on the customers. There's an inventory database, there's a finance database, there's a procurement database, but all of that needs to connect, and that's a challenge.
For marketers, what that really means is Customer A, how are they interacting with the firm? When did they first start buying? What are the installations? Which salespeople did they talk to? Having a foundation of that kind of first-party customer-focused data is important.
Then we talk about letting a thousand flowers bloom. Trying different things but looking at one aspect of customer engagement, either acquisition, retention, growth, or advocacy. You're really trying different things, looking at ROI and seeing where the biggest bang for your buck is.
Once you learn what was working for you, then you're slowly expanding into other aspects of customer engagement. Eventually, you will reach a place where so far you've used vendors and off-the-shelf products and you have to really invest, and you get into a board level position of investing in your own data science team, building it in-house or buying an AI company.
The last stage is very fascinating, which we saw. Specifically for AI companies, they have started taking all of these capabilities they've built in-house and they've turned around and built it into a platform that they are now selling AI as a service for other companies and building a new revenue stream. That is something which was really interesting to see how every company who reached this transformation has turned around and done a services platform.
Kerry: What industries then lend themselves best to this, which can benefit the most from this? It definitely seems like data can benefit about any industry, but when I'm hearing you talk about services, that seems like an area really rich to benefit from AI.
Raj: Absolutely. Services definitely. I think data services especially. In the book, we talk about Washington Post and Starbucks in detail in terms of what they did.
I think Caterpillar is definitely getting into that world in terms of its ability to fit the tractors with cameras that can understand and detect which crops need pesticide and can target the crops in terms of pesticide spraying. One is you can see it as a big savings for the farmers, but it's also less pesticides on crops and that's a good thing. Then you're building on top of that now other capabilities of how you can operate better, algorithms for big farms, how to optimize the way you use your spraying capability. These things can become services that they can then provide big industrial farms.
In another example, there's a company where they have this IoT device that's a microphone, where people use it in oil rigs and really hazardous situations. Where it is really helpful, first, is it was done to ensure somebody can know where they are and they can communicate with each other. But what this company found is that based how these microphones are and people are hovering on their movements, you can see if some issue is really a dangerous situation or not, and whether you should react or not to actually then improve the safety of your employees in the oil rigs. Once they found that this data can be used by their clients, they're building a platform which then provides a data service on top of the microphones for improving worker safety.
Kerry: It's possible there are some people who have heard of the Internet-of-Things but don't know what it is. Internet connected things that are not computers, basically. Right?
Kerry: How far does that go and how weirded out do people get? I would be a little embarrassed if I was flipping the wrong switch on a thing 167 times. I guess it's valuable data for a product developer to know that I can't figure it out.
Raj: I think that what you're saying is absolutely true. I think that is important information from the perspective of product developers to understand that. It's not just one Kerry doing it, if there are a lot of customers who are doing that, then that is certainly something your product developers need to pay attention to. That's the beauty of what I call extending the purchasing funnel.
So far, we've only looked at awareness, interest, desire, and action, before the internet and before data and technology. Basically, the purchase funnel stops at action, when the customers buy the product. Then you don't know how they're using it. You just wait for them to come back and buy the next product.
Once you know about how consumption is happening, how people are using the product, then it's an ongoing relationship of you know what needs to be fixed, what new things people want, where they spend the most time, what they find valuable. All of this insight is what is really allowing for new innovations to happen.
Kerry: Where are the missed opportunities? Are there areas now that are still relatively untapped that people could get into and be the early adopters and really benefit from that first-mover advantage?
Raj: Great question. Right now, I feel like what differentiates companies from doing this better than the others, the differentiation I feel is in the culture and the process of the company. I think data science as a service is getting commoditized, but the differentiation is how much you've invested in knowing about your customers, how much you've invested in collecting the data, and how much your process of marketing decision-making and marketing strategy is attuned to getting these customer insights and implementing personalized marketing actions based on those insights.
Kerry: If you were to talk to someone who is really just starting out, they have a brand new business and they can build their systems and their technology stack any way they want, what would you tell them, to take the best advantage of AI what should they do?
Raj: Great question. I think the first thing I would say is good for you, because you don't have legacy issues. You are starting in a place where you can build it as you want, and you can really think about customer-first. I think that is number one.
You can really start thinking about omnichannel. I think that's the first thing I would begin with. If you are from scratch building your systems, think about serving your customers across channels and being able to see them across those channels, because customers will engage with you in all the places and you need to be able to see them. The technology system that allows you to do that, whether they're online or in the store, or talking to your salesperson, you're able to track it. Even in a tradeshow, there's a good system for doing all of that.
The second thing I would say is almost everybody is a web 2.0, web 3.0 company now, everybody has a website, everybody has either a LinkedIn or a forum. Tradeshows are all virtual now, and maybe there's a hybrid tradeshow that will happen in the future. You are tracking, even if you're a startup, all of your interactions with your customers. You will have Google Analytics, or you may have HubSpot or Adobe in the back end, giving you data on how your customers are engaging with you. The same vendors will also be happy to give you reports on any kind of analytics you want, basic analytics.
That's where I would begin. Once I collect the data, I would begin with customizing an email campaign, or customizing your sales call plan, or customizing how your salesperson approaches your customers, arming your salesforce with information about your customers, and see where that goes.
Kerry: You mentioned a couple of times that data is the raw fuel for AI. How can people keep their data clean? I know that's one of the biggest problems. Sales isn't paid to input data, they're paid to make sales, so they're just getting it in as fast as they can. How do you keep your data clean?
Raj: We talk about 80% of the work is data and 20% is actually the estimation of the models, and that is true. The digitization helps, but there are always places where there is human input, that's where there are challenges. But human input is necessary because that also gives you qualitative nuanced information that can be a differentiator.
It is something you have to struggle with and you cannot ignore. I wouldn't say only use data that is completely coming from digital sources, because the real rich information is actually in people's minds and that's what you want to tap into.
Kerry: How deep an understanding of the technology does a marketer need to have? I think they're a little scared. We can get a little scared.
Raj: Absolutely. That's a great question. This is something that is so relevant. I say in my class when I teach marketing analytics and digital marketing that I want you to be smart consumers of analytics. What I mean by that is it's not like you need to go and program and learn Python or R or any other new thing that comes up. You need to understand what data can do and what is possible.
I think the biggest skill marketers can have is being open-minded and understanding about the power of data. And I think marketers are there and I think a lot of marketers understand it.
I think the ones I see are successful are people who are collaborators and are people who are able to work across different functions. Marketers who are really successful form really good relationships with the technology folks, with the data science folks, with the operations folks, with the finance folks. The way I think about marketing, I think marketers need to think of themselves as the chief advocate of customers within an organization. They are the ones who are talking about and really focused on doing what is right for the customers across the organization.
If an organization recognizes the marketer's role like that and if the marketer themselves recognizes that's their role, they will see that for them to be better, to be good at their job, they need the help across all divisions. That's what we talk about in the book, also, in the expansion stage when we talk about the marketing AI champion. You're really looking at the skills of a person who is a connector and a collaborator. Being able to put together teams and being able to understand and harness the knowledge of a team, I think, is the most important skill marketers need, more than coding or analytics.
Kerry: Raj, where can people learn more and where can they get their copy of The AI Marketing Canvas?
Raj: The AI Marketing Canvas is on Amazon. We would love for you to go check it out, give us your feedback, and send us your reviews. Jim and I are on LinkedIn. We'd love to hear from you. We continue a conversation there about new topics and this is an ongoing effort. This is just the beginning for marketing, so there are going to be more and more fabulous stories that are going to come about how brands have used AI.
Kerry: And there's a website for the book as well?
Raj: There is a website, thank you, at AIMCbook.com.
Kerry: Great. Thank you so much for joining today. I learned a lot. I hope everybody buys the book through your site so that you can get all that rich data on what they do with it.
Raj: Absolutely. That's right. Thank you, Kerry. Thank you for having me. It has been a pleasure.
Kerry: Thank you for listening here to the very end. This has been the Marketing Smarts Podcast. Talk with you again soon.
Published on August 19, 2021
Professor Rajkumar Venkatesan, co-author of the book The AI Marketing Canvas: A Five-Stage Road Map to Implementing Artificial Intelligence in Marketing. He teaches Marketing Technology Products, Marketing Strategy, and Marketing Analytics at the UVA Darden School of Business. His research focuses on analytics as it relates to marketing return on investment, customer lifetime value, mobile marketing, and the global political economy.
You may like these other MarketingProfs articles related to Metrics & ROI: